[8] Agassiz’s description of the Pterichthys, as quoted by Humboldt, in his Cosmos.
[9] From Murchison’s Silurian System.
[10] These scales, which occur in a detached state, in a stratified clay of the Old Red Sandstone, near Cromarty, present for their size a larger extent of cover than the scales of any other Ganoid.
[11] A peculiarity which also occurs in the anterior dorsal of the Dipterus.
[12] From the head of Raja clavata.
[13] The darker, upper patch in this figure indicates a portion in which the scales of the fins in the fossil still retain their enamel;—the lighter, a portion from which the enamel has disappeared.
[14] The Acanths of the Coal Measures possess the cranial buckler.
[15] Professor Owen, in fixing the homologies of the ichthyic head, differs considerably from Cuvier; but his view seems to be demonstrably the correct one. It will, however, be seen, that in my attempted comparison of the divisions of the ancient ganoid cranium with those of the craniums of existing fishes, the points at issue between the two great naturalists are not involved, otherwise than as mere questions of words. The matter to be determined, for instance, is not whether plate A in the skulls of the cod and Coccosteus be the homologue of a part of the occipital or that of a part of the parietal bones, but whether plate A in the Coccosteus be the homologue of plate A in the cod. The letters employed I have borrowed from Agassiz’s restoration of the Coccosteus; whereas the figures intimate divisions which the imperfect keeping of the specimens on which the ichthyologist founded did not enable him to detect.
[16] The jaws (10, 10) which exhibit in the print their greatest breadth, would have presented in the animal, seen from beneath, their narrow under-edges, and have nearly fallen into the line of the sub-opercular plates, (13, 13.)
[17] In all probability it is likewise the principle of the placoid skull. The numerous osseous points by which the latter is encrusted, each capable of increase at the edges, seem the minute bricks of an ample dome. It is possible, however, that new points may be formed in the interstices between the first formed ones, as what anatomists term the triquetra or Wormiana form between the serrated edges of the lambdoidal suture in the human skull; and that the osseous surface of the cerebral dome may thus extend, as the dome itself increases in size, not through the growth of the previously existing pieces,—the minute bricks of my illustration,—but through the addition of new ones. Equally, in either case, however, that essential difference between the placoid skull and the placoid vertebra, to which I have referred, appears to hinge on the circumstance, that while the osseous nucleus of each vertebral centrum could form, in even its most complicated shape, from a single point, the osseous walls of the cranium had to be formed from hundreds. The accompanying diagram serves to show after what manner the vertebral centrum in the Ray enlarges with the growth of the animal, by addition of bony matter external to the point in the middle, at which ossification first begins. The horizontal lines indicate the lines of increment in the two internal cones which each centrum comprises, and the vertical ones the lines of increment in the lateral pillars.